A method for identifying the relationship between surface electromyography (EMG) signals and implanted EMG signals.

Abstract

This application discloses a method for identifying the relationship between surface electromyography (EMG) signals and implanted EMG signals. The method includes: acquiring first feature information of the surface EMG signal, second feature information of the implanted EMG signal, and third feature information of the myophone signal; calculating the correlation coefficient between the first and second feature information; determining whether a mapping relationship exists between the surface EMG signal and the implanted EMG signal based on the correlation coefficient; verifying that the third feature information of the myophone signal has an information compensation effect on the first feature information of the surface EMG signal when a mapping relationship exists; and optimizing the mapping relationship between the surface EMG signal and the implanted EMG signal based on the information compensation effect. This application can identify whether a significant mapping relationship exists between surface EMG signals and implanted EMG signals and improves the recognition accuracy of the mapping relationship between surface EMG signals and implanted EMG signals, and can be widely applied in the field of EMG signal processing technology.

Description

Technical Field

[0001] This application relates to the field of electromyography signal processing technology, and in particular to a method for identifying the relationship between surface electromyography signals and implanted electromyography signals. Background Technology

[0002] In clinical and basic research, electromyography (EMG) monitoring technology is mainly divided into two categories: surface EMG monitoring and implantable EMG monitoring. sEMG collects signals through electrodes on the skin surface, offering advantages such as being non-invasive and easy to operate. However, its signals are susceptible to attenuation by subcutaneous tissue and noise interference during transmission, resulting in low spatial resolution and difficulty in accurately distinguishing specific muscle activities. EMG, on the other hand, directly collects signals from within the muscle through implanted electrodes, offering high signal-to-noise ratio and excellent resolution. However, it suffers from drawbacks such as high invasiveness and limited implantation time, and cannot achieve long-term continuous monitoring. Therefore, establishing a reliable mapping relationship between the two is of great significance for promoting the development of long-term non-invasive EMG monitoring.

[0003] In existing technologies, related studies mostly rely on simple statistical correlation analysis to analyze the relationship between surface electromyography (EMG) signals and implanted EMG signals. This has failed to fully verify and integrate myophone signal (MMG), a complementary signal source that can provide information on muscle mechanical activity. Because the value of MMG signals has not been fully explored, it is difficult to make up for the lack of information in surface EMG signals. As a result, the accuracy of the mapping relationship between surface EMG signals and implanted EMG signals is insufficient, and it is impossible to provide a reliable scientific basis for long-term non-invasive EMG monitoring. Summary of the Invention

[0004] The main objective of this application is to propose a method for identifying the relationship between surface electromyography (EMG) signals and implanted EMG signals. This method can identify whether there is a significant mapping relationship between surface EMG signals and implanted EMG signals, and improves the accuracy of identifying the mapping relationship between surface EMG signals and implanted EMG signals, making the identified mapping relationship more reliable.

[0005] To achieve the above objectives, one aspect of this application proposes a method for identifying the relationship between surface electromyography (EMG) signals and implanted EMG signals, including: Acquire surface electromyographic signals, implanted electromyographic signals, and myophone signals of the target muscle; Acquire the first feature information of the surface electromyography signal, the second feature information of the implanted electromyography signal, and the third feature information of the myocardial sound signal; Calculate the correlation coefficient between the first feature information and the second feature information; Based on the correlation coefficient, determine whether there is a mapping relationship between the surface electromyography signal and the implanted electromyography signal; In the presence of the aforementioned mapping relationship, it is verified that the third feature information of the myophone signal has an information compensation effect on the first feature information of the surface electromyography signal; Based on the information compensation effect, the mapping relationship between the surface electromyography signal and the implanted electromyography signal is optimized.

[0006] In some embodiments, the first feature information, the second feature information, and the third feature information include time-domain features, frequency-domain features, time-frequency features, and nonlinear dynamic system features; The steps for obtaining the time-domain features, frequency-domain features, time-frequency features, and nonlinear dynamic system features include: The envelope information corresponding to the surface electromyography signal, implanted electromyography signal and myophone signal is obtained respectively, and the time-domain features in the first feature information, the time-domain features in the second feature information and the time-domain features in the third feature information are extracted from each corresponding envelope information. Calculate the power spectra corresponding to the surface electromyography signal, implanted electromyography signal, and myophone signal respectively, and extract the frequency domain features in the first feature information, the frequency domain features in the second feature information, and the frequency domain features in the third feature information from each corresponding power spectrum; Wavelet transforms are performed on the surface electromyography signal, implanted electromyography signal, and myophone signal, respectively. The time-frequency features in the first feature information, the time-frequency features in the second feature information, and the time-frequency features in the third feature information are extracted from the results of the corresponding wavelet transforms. The Lyapunov exponent, correlation dimension, entropy, information entropy, and complexity corresponding to surface electromyography (EMG), implanted EMG, and myosal signals are calculated respectively to obtain the nonlinear dynamic system features in the first feature information, the nonlinear dynamic system features in the second feature information, and the nonlinear dynamic system features in the third feature information.

[0007] In some embodiments, calculating the correlation coefficient between the first feature information and the second feature information includes: Calculate the covariance between the first feature information and the second feature information; Determine the standard deviation of the first feature information and the standard deviation of the second feature information; The correlation coefficient is calculated based on the covariance, the standard deviation of the first feature information, and the standard deviation of the second feature information.

[0008] In some embodiments, determining whether a mapping relationship exists between the surface electromyography (EMG) signal and the implanted EMG signal based on the correlation coefficient includes: Determine the statistical significance test value between the first feature information and the second feature information, wherein the statistical significance test value is used to characterize the probability value that the difference between the first feature information and the second feature information is caused by random factors; If the absolute value of the correlation coefficient is greater than a first preset threshold and the statistical significance test value is less than a second preset threshold, it is determined that there is a mapping relationship between the surface electromyography signal and the implanted electromyography signal.

[0009] In some embodiments, the verification of the third feature information of the myophone signal has an information compensation effect on the first feature information of the surface electromyography signal, including: Calculate the difference between the first feature information and the third feature information; The first feature information, the second feature information, and the third feature information are used to construct a feature matrix. Principal component analysis is then performed on the feature matrix to obtain the analysis results. Based on the differences and the analysis results, it is verified that the third characteristic information of the myophone signal has an information compensation effect on the first characteristic information of the surface electromyography signal. In some embodiments, performing principal component analysis on the feature matrix to obtain analysis results includes: Calculate the covariance matrix of the feature matrix, and perform eigenvalue decomposition on the covariance matrix to obtain each eigenvalue and the corresponding eigenvector; Based on the eigenvalues, select k eigenvectors corresponding to the eigenvalues ​​to form the principal component space, where K is a positive integer greater than zero; The feature matrix is ​​projected onto the principal component space to obtain the dimensionality-reduced data matrix; Calculate the first spatial distance between the first feature information and the second feature information, and the second spatial distance between the first feature information and the third feature information in the dimensionality-reduced data matrix; Principal component loading information is obtained based on the feature vector; The contribution rate of each principal component is calculated based on the aforementioned eigenvalues; The first spatial distance, the second spatial distance, the contribution rate, and the load information are used as the analysis results.

[0010] In some embodiments, verifying that the third feature information of the myophone signal has an information compensation effect on the first feature information of the surface electromyography signal based on the difference and the analysis result includes: If the difference meets the preset conditions, it is determined that the first condition for verification is met; If the second spatial distance is greater than the first spatial distance, it is determined that the second verification condition is met; Based on the contribution rate and the load information, one or more principal components dominated by the third feature information are identified, and the total contribution rate of the one or more principal components is calculated. If the total contribution rate is greater than a preset contribution rate threshold, the third condition for verification is determined to be met. If the first condition, the second condition, and the third condition are all satisfied, it is determined that the third feature information of the myophone signal has an information compensation effect on the first feature information of the surface electromyography signal. To achieve the above objectives, another aspect of this application proposes a device for identifying the relationship between surface electromyographic signals and implanted electromyographic signals, the device comprising: The signal acquisition module is used to acquire surface electromyographic signals, implanted electromyographic signals, and myophone signals of the target muscle. The feature information acquisition module is used to acquire the first feature information of the surface electromyography signal, the second feature information of the implanted electromyography signal, and the third feature information of the myophone signal; The correlation coefficient calculation module is used to calculate the correlation coefficient between the first feature information and the second feature information. The relationship determination module is used to determine, based on the correlation coefficient, whether there is a mapping relationship between the surface electromyography signal and the implanted electromyography signal; The information compensation effect verification module is used to verify that, in the presence of the mapping relationship, the third feature information of the myophone signal has an information compensation effect on the first feature information of the surface electromyography signal. The relationship optimization module is used to optimize the mapping relationship between the surface electromyography signal and the implanted electromyography signal based on the information compensation effect.

[0011] To achieve the above objectives, another aspect of this application provides an electronic device, which includes a memory and a processor. The memory stores a computer program, and the processor executes the computer program to implement the method described above.

[0012] To achieve the above objectives, another aspect of the embodiments of this application proposes a computer-readable storage medium storing a computer program that, when executed by a processor, implements the methods described above.

[0013] This application also discloses a computer program product or computer program, which includes computer instructions stored in a computer-readable storage medium. A processor of a computer device can read the computer instructions from the computer-readable storage medium and execute the computer instructions, causing the computer device to perform the aforementioned method.

[0014] The embodiments of this application include at least the following beneficial effects: This application provides a method for identifying the relationship between surface electromyography (EMG) signals and implanted EMG signals. This method first acquires surface EMG signals, implanted EMG signals, and myophone signals of the target muscle; acquires first feature information of the surface EMG signals, second feature information of the implanted EMG signals, and third feature information of the myophone signals; calculates the correlation coefficient between the first and second feature information; and based on the correlation coefficient, determines whether a mapping relationship exists between the surface EMG signals and the implanted EMG signals, thus identifying whether a significant mapping exists between them. The study aims to objectively and quantitatively determine the linear correlation between surface electromyography (EMG) signals and implanted EMG signals. In cases where a mapping relationship exists, it verifies that the third feature information of the myophone signal compensates for the first feature information of the surface EMG signal, fully exploring the effective feature dimensions in the myophone signal that can compensate for missing information in the surface EMG signal. Finally, based on this information compensation effect, the mapping relationship between surface EMG signals and implanted EMG signals is optimized, significantly improving the recognition accuracy of the mapping relationship and making the identified mapping relationship more reliable. This provides reliable technical support for achieving long-term non-invasive EMG monitoring. Attached Figure Description

[0015] Figure 1 This is a schematic diagram of an implementation environment provided in an embodiment of this application; Figure 2 This is a flowchart illustrating a method for identifying the relationship between surface electromyography (EMG) signals and implanted EMG signals, as provided in an embodiment of this application. Figure 3 A schematic diagram illustrating the temporal characteristics of surface electromyography (EMG), implanted EMG, and myophone signals provided in the embodiments of this application; Figure 4 A schematic diagram illustrating the time-frequency characteristics of surface electromyography (EMG), implanted EMG, and myophone signals provided in the embodiments of this application; Figure 5 This is a flowchart illustrating the calculation of the correlation coefficient between the first feature information and the second feature information, provided in an embodiment of this application. Figure 6 This is a flowchart provided in an embodiment of the present application to verify that the third feature information of the myophone signal has an information compensation effect on the first feature information of the surface electromyography signal; Figure 7 This is a schematic diagram of the structure of a device for identifying the relationship between surface electromyography (EMG) signals and implanted EMG signals provided in an embodiment of this application; Figure 8 This is a schematic diagram of the hardware structure of the electronic device provided in the embodiments of this application. Detailed Implementation

[0016] To make the objectives, technical solutions, and advantages of this application clearer, the following detailed description is provided in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative of this application and are not intended to limit it. In the following description, when referring to the accompanying drawings, unless otherwise indicated, the same numbers in different drawings represent the same or similar elements. The embodiments described in the following exemplary embodiments do not represent all embodiments consistent with those of this application; they are merely examples of apparatuses and methods consistent with some aspects of the embodiments of this application as detailed in the appended claims.

[0017] It is understood that the terms "first," "second," "third," "fourth," etc. (if present) in the specification and accompanying drawings of this application are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that such data can be interchanged where appropriate so that the embodiments of this application described herein can be implemented in orders other than those illustrated or described herein. Furthermore, the terms "comprising" and "having," and any variations thereof, are intended to cover non-exclusive inclusion; for example, a process, method, system, product, or apparatus that comprises a series of steps or units is not necessarily limited to those steps or units explicitly listed, but may include other steps or units not explicitly listed or inherent to such processes, methods, products, or apparatus.

[0018] It should be understood that in this application, "at least one (item)" means one or more, and "more than" means two or more. "And / or" is used to describe the relationship between related objects, indicating that three relationships can exist. For example, "A and / or B" can represent three cases: only A exists, only B exists, and both A and B exist simultaneously, where A and B can be singular or plural. The character " / " generally indicates that the preceding and following related objects are in an "or" relationship. "At least one (item) of the following" or similar expressions refer to any combination of these items, including any combination of single or plural items. For example, at least one (item) of a, b, or c can represent: a, b, c, "a and b", "a and c", "b and c", or "a and b and c", where a, b, and c can be single or multiple.

[0019] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this application belongs. The terminology used herein is for the purpose of describing embodiments of this application only and is not intended to limit this application.

[0020] In existing technologies, related studies mostly rely on simple statistical correlation analysis to analyze the relationship between surface electromyography (EMG) signals and implanted EMG signals. This has failed to fully verify and integrate myophone signal (MMG), a complementary signal source that can provide information on muscle mechanical activity. Because the value of MMG signals has not been fully explored, it is difficult to make up for the lack of information in surface EMG signals. As a result, the accuracy of the mapping relationship between surface EMG signals and implanted EMG signals is insufficient, and it is impossible to provide a reliable scientific basis for long-term non-invasive EMG monitoring.

[0021] In view of this, this application provides a method for identifying the relationship between surface electromyography (EMG) signals and implanted EMG signals. This scheme first acquires surface electromyography (EMG) signals, implanted EMG signals, and myophone signals of the target muscle. It then acquires the first feature information of the surface EMG signal, the second feature information of the implanted EMG signal, and the third feature information of the myophone signal. The correlation coefficient between the first and second feature information is calculated. Based on the correlation coefficient, it determines whether a mapping relationship exists between the surface EMG signal and the implanted EMG signal, identifying a significant mapping relationship and objectively quantifying the degree of linear correlation between them. In cases where a mapping relationship exists, it verifies that the third feature information of the myophone signal compensates for the first feature information of the surface EMG signal, fully exploring the effective feature dimensions in the myophone signal that can compensate for missing information in the surface EMG signal. Finally, based on the information compensation effect, the mapping relationship between the surface EMG signal and the implanted EMG signal is optimized, significantly improving the recognition accuracy of the mapping relationship and making the identified mapping relationship more reliable, providing reliable technical support for long-term non-invasive EMG monitoring. The method for identifying the relationship between surface electromyography (EMG) signals and implanted EMG signals provided in this application relates to the field of EMG signal processing technology. This method can be applied to a terminal, a server, or software running on either a terminal or a server. In some embodiments, the terminal can be a smartphone, tablet, laptop, desktop computer, smart speaker, smartwatch, or in-vehicle terminal, but is not limited to these. The server can be configured as an independent physical server, a server cluster or distributed system composed of multiple physical servers, or a cloud server providing basic cloud computing services such as cloud services, cloud databases, cloud computing, cloud functions, cloud storage, network services, cloud communication, middleware services, domain name services, security services, CDN, and big data and artificial intelligence platforms. The server can also be a node server in a blockchain network. The software can be an application that implements the method for identifying the relationship between surface EMG signals and implanted EMG signals, but is not limited to the above forms.

[0022] This application can be used in a wide variety of general-purpose or special-purpose computer system environments or configurations. Examples include: personal computers, server computers, handheld or portable devices, tablet devices, multiprocessor systems, microprocessor-based systems, set-top boxes, programmable consumer electronics devices, network PCs, minicomputers, mainframe computers, and distributed computing environments including any of the above systems or devices. This application can be described in the general context of computer-executable instructions executed by a computer, such as program modules. Generally, program modules include routines, programs, objects, components, data structures, etc., that perform specific tasks or implement specific abstract data types. This application can also be practiced in distributed computing environments where tasks are performed by remote processing devices connected via a communication network. In distributed computing environments, program modules can reside in local and remote computer storage media, including storage devices.

[0023] like Figure 1 The diagram shown is a schematic representation of an implementation environment provided in an embodiment of this application. (Refer to...) Figure 1 The implementation environment includes at least one terminal 102 and a server 101. The terminal 102 and the server 101 can be connected via a network, either wirelessly or via a wired connection, to complete data transmission and exchange.

[0024] Server 101 can be a standalone physical server, a server cluster or distributed system consisting of multiple physical servers, or a cloud server that provides basic cloud computing services such as cloud services, cloud databases, cloud computing, cloud functions, cloud storage, network services, cloud communication, middleware services, domain name services, security services, CDN (Content Delivery Network), and big data and artificial intelligence platforms.

[0025] Additionally, server 101 can also be a node server in a blockchain network. Blockchain is a novel application model of computer technologies such as distributed data storage, peer-to-peer transmission, consensus mechanisms, and encryption algorithms.

[0026] Terminal 102 can be a smartphone, tablet, laptop, desktop computer, smart speaker, smartwatch, etc. It can also be a vehicle-mounted terminal of the various device types described above, but is not limited to these. Terminal 102 and server 101 can be directly or indirectly connected via wired or wireless communication, and this embodiment does not impose any limitations.

[0027] Exemplary based on Figure 1 The implementation environment shown in this application embodiment provides a method for identifying the relationship between surface electromyography (EMG) signals and implanted EMG signals. The following description uses the application of this method in server 101 as an example. It is understood that this method can also be applied to terminal 102.

[0028] Reference Figure 2 , Figure 2 This is a flowchart illustrating a method for identifying the relationship between surface electromyography (EMG) signals and implanted EMG signals applied to a server, provided in an embodiment of this application. The executing entity of this method can be any of the aforementioned computer devices (including a server or terminal). (Refer to...) Figure 2 The method may include the following steps: S100: Acquire surface electromyographic signals, implanted electromyographic signals, and myosal signals of the target muscle.

[0029] For example, embodiments of this application can acquire surface electromyography (EMG) signals, implanted EMG signals, and myocardial sound signals of a target muscle using a multi-channel synchronous EMG acquisition system. In the multi-channel synchronous EMG acquisition system, surface EMG signals of the target muscle are acquired using a dry electrode wristband, implanted EMG signals of the target muscle are acquired using an implanted flexible electrode wire, and myocardial sound signals of the target muscle are acquired using a stretchable liquid metal sensor. The multi-channel synchronous EMG acquisition system can ensure the alignment of EMG signals at different levels on the time scale through a unified sampling clock and triggering mechanism.

[0030] Furthermore, after acquiring the surface electromyography (EMG), implanted EMG, and myophone signals of the target muscle, the process includes preprocessing these signals. Preprocessing includes filtering, power frequency interference removal, and artifact removal to reduce the influence of EMG, eye movement, motion artifacts, and device noise. Then, the surface EMG, implanted EMG, and myophone signals are segmented by a preset time length to obtain multiple time segments. Each channel undergoes amplitude normalization and time resampling to construct a uniform-length, multi-channel, multi-level signal time sequence.

[0031] S200: Acquire the first feature information of the surface electromyography signal, the second feature information of the implanted electromyography signal, and the third feature information of the myophone signal.

[0032] The first feature information, the second feature information, and the third feature information all include time-domain features, frequency-domain features, time-frequency features, and nonlinear dynamic system features. For example, the steps for obtaining the time-domain features, frequency-domain features, time-frequency features, and nonlinear dynamic system features include S210-S240: S210. Obtain the envelope information corresponding to the surface electromyography signal, implanted electromyography signal and myosal signal respectively, and extract the time domain features in the first feature information, the time domain features in the second feature information and the time domain features in the third feature information from each corresponding envelope information.

[0033] For example, firstly, the surface electromyography (EMG) signal, implanted EMG signal, and myophone signal are subjected to bandpass filtering. This processing uses a Butterworth filter of a specified order (e.g., 3rd order). By setting a low cutoff frequency and a high cutoff frequency, a bandpass filtering condition is formed to retain the effective EMG signal frequency band, thereby improving the signal-to-noise ratio. Then, the bandpass-filtered surface EMG signal, implanted EMG signal, and myophone signal are subjected to notch filtering. This notch filtering uses a notch filter designed for a specific interference frequency (e.g., 50Hz power frequency and its harmonics). By setting the corresponding quality factor parameter, deep attenuation is generated within the specific frequency and its adjacent narrow band range, thereby effectively filtering out continuous periodic interference at that frequency point.

[0034] Furthermore, a Hilbert transform is performed on the surface electromyography (EMG), implanted EMG, and myophone signals after the above filtering process to obtain an analytical signal. The absolute value of the analytical signal is taken as the original envelope information (reflecting the amplitude variation trend of the signal). Gaussian smoothing filtering (based on Gaussian smoothing parameters) is applied to the envelope information to obtain a smoothed envelope to reduce noise interference. Next, the mean (as baseline) and standard deviation of the surface EMG, implanted EMG, and myophone signals are calculated respectively. Combined with a predetermined threshold factor, a judgment threshold for distinguishing effective EMG activity from background noise is calculated according to the formula "threshold = baseline + threshold factor × standard deviation". Continuous regions in the smoothed envelope that exceed the above threshold (i.e., regions with envelope value > threshold) are identified, and regions with durations shorter than the minimum duration (converted to the number of sampling points) are excluded (filtering out spurious signals caused by transient noise). For the identified continuous regions, the peak values ​​of surface electromyography (EMG), implanted EMG, and myophone signals within those regions are calculated, and regions with peak values ​​below the envelope peak threshold are excluded (filtering out invalid noise signals with excessively low amplitude and valid signal segments with excessively low signal-to-noise ratio). Finally, the intervals between the remaining regions are calculated (converted to the number of sampling points). If the interval is less than the minimum interval, the two regions are merged (to avoid segmenting the signal of the same action due to brief noise interruptions). The duration of each merged region is calculated (converted to the number of sampling points), and regions with durations exceeding the maximum duration are excluded (filtering out abnormally long invalid signals). After the above processing, the envelope information corresponding to the valid signals of surface EMG, implanted EMG, and myophone is obtained. Features such as absolute value integral (reflecting signal strength), average energy, variance, standard deviation (essentially the average energy of the signal), number of zero crossings (or number of threshold crossings, essentially the rate of signal change), and first derivative curve are extracted from the envelope information corresponding to the valid signals as time-domain features, such as... Figure 3 As shown, Figure 3 This is a schematic diagram illustrating the time-domain characteristics of surface electromyography (EMG), implanted EMG, and myophone signals provided in the embodiments of this application. Figure 3 The paper presents the temporal characteristics of four channels of surface electromyography (EMG) signals, one channel of implanted EMG signals, and one channel of myophone signals.

[0035] S220. Calculate the power spectra corresponding to the surface electromyography signal, implanted electromyography signal and myophone signal respectively, and extract the frequency domain features in the first feature information, the frequency domain features in the second feature information and the frequency domain features in the third feature information from each corresponding power spectrum.

[0036] For example, in the embodiments of this application, the power spectra of the Fourier transforms corresponding to surface electromyography signals, implanted electromyography signals and myophone signals are calculated by periodograms, and features such as the frequency corresponding to peak power, the frequency corresponding to average power (MPF), and the frequency corresponding to median power (MF) are extracted from each corresponding power spectrum as frequency domain features.

[0037] S230. Perform wavelet transform on the surface electromyography signal, implanted electromyography signal and myophone signal respectively, and extract the time-frequency features in the first feature information, the time-frequency features in the second feature information and the time-frequency features in the third feature information from the results of each corresponding wavelet transform.

[0038] For example, wavelet transform is used to analyze the time-frequency characteristics of a signal, and features such as wavelet energy entropy are extracted as time-frequency features, such as... Figure 4 As shown, Figure 4 This is a schematic diagram illustrating the time-frequency characteristics of surface electromyography (EMG), implanted EMG, and myophone signals provided in the embodiments of this application. Figure 4 The paper presents the time-frequency characteristics of surface electromyography (EMG) signals from four channels, the time-frequency characteristics of implanted EMG signals from one channel, and the time-frequency characteristics of myophone signals from one channel.

[0039] S240. Calculate the Lyapunov exponent, correlation dimension, entropy, information entropy, and complexity corresponding to the surface electromyography signal, implanted electromyography signal, and myophone signal, respectively, to obtain the nonlinear dynamic system features in the first feature information, the nonlinear dynamic system features in the second feature information, and the nonlinear dynamic system features in the third feature information.

[0040] For example, in an embodiment of this application, the step of calculating the Lyapunov index includes: reconstructing the phase spaces of the electromyography (EMG) signal, implanted EMG signal, and myophone signal based on a preset embedding dimension m and a time delay τ, respectively, with the following formula: Where i = 1, 2, …, N-(m-1)τ in, The original signal is N, and the signal length is N.

[0041] Specifically, the embedding dimension can be determined using the spurious nearest neighbor method. For implanted electromyography (EMG) signals and surface EMG signals, m is typically taken as 5~8, while for myosalpingography (MIG) signals, m is taken as 3~5. The time delay τ can be determined using the autocorrelation function or mutual information method. Further, for each point... Find its nearest neighbor in the reconstructed phase space. Ensure initial distance Less than the threshold, and To avoid using spurious nearest neighbors, track the evolution of each pair of neighboring points in the phase space and calculate the distance after k steps. .when When the distance exceeds a threshold (typically 1 / 3 of the attractor diameter), the separation distance is recorded and the nearest neighbor is searched again. Finally, the maximum Lyapunov exponent (MLE) is calculated based on the initial distance and the distance after k steps.

[0042] The steps for calculating the correlation dimension include: First, calculating the correlation integral, the formula of which is:

[0043] Where H is the Heaviside function, which is 1 when the parameter is greater than 0, and 0 otherwise; r is the distance threshold. Euclidean distance Secondly, the correlation integral is plotted on log-log coordinates to identify the linear region (scaling region). Finally, the slope of the scaling region is used to estimate the correlation dimension. Its formula is:

[0044] In this embodiment of the application, by repeatedly calculating for different embedding dimensions m, when... When the relationship stabilizes, the corresponding value is the correlation dimension.

[0045] The steps for calculating entropy include: setting the template length m and the similarity tolerance r; constructing a vector sequence based on the template length. Where i = 1, 2, ..., N-m+1; calculate the distance between any two vectors in the vector sequence, the formula is: ; Statistical satisfaction number of vector pairs ,exclude The formula for calculating the number of vector pairs is: Finally, increase the template length by 1 (i.e., m+1) and repeat the above process to obtain... Based on its calculation of the entropy of surface electromyography (EMG), implanted EMG, or myosalpingography (MIG) signals, the formula is as follows: .

[0046] The steps for calculating information entropy include: dividing the time series into K intervals and calculating the probability of each interval. Then, the probability is obtained using histogram or kernel density estimation methods. The probability distribution is solved using kernel density estimation (KDE) because it is a smoother and more accurate probability distribution estimation method than histograms, and it does not depend on fixed interval divisions. Kernel density analysis formula:

[0047] in, This is the estimated probability density at point x. For the sample size, For the first One sample point, The bandwidth parameter controls the smoothness. For kernel functions, satisfying and In the electromyography signal analysis of this application, a Gaussian kernel was selected: It is infinitely differentiable and robust to outliers.

[0048] This application embodiment considers the limited dataset size and employs leave-one-out cross-validation to obtain the accurate optimal bandwidth. The core idea of ​​cross-validation (LOOCV) is to minimize the error between the kernel density estimate and the true distribution, and its objective function is:

[0049] in, For the bandwidth parameters to be optimized, For KDE estimation using all data and bandwidth, To estimate the density of points using the remaining samples after excluding the samples, The total number of samples, To estimate the smoothness measure of the distribution, This is used as a measure of model fit. Among them, the optimal bandwidth... The calculation formula is: ; Calculate discrete probability mass Among them, the distance between evaluation points for m is the number of evaluation points. To determine the coefficients for the range of evaluation points ( The molecular reaction is the evaluation point range. According to discrete probability mass The formula for calculating information entropy is:

[0050] The steps for calculating complexity include: First, converting the electromyographic (EMG) signal, implanted EMG signal, and myophone signal into binary sequences, respectively, using the following formula: if ,otherwise ,in, Let the mean of the signal be denoted as . Next, the binary sequence is divided into different subsequences, and the minimum number of replication steps required to generate a new subsequence is calculated, i.e., the complexity count. The complexity is calculated based on complexity counting and sequence length. Its formula is: .

[0051] Furthermore, embodiments of this application also include calculating sample entropy, the steps of which include: establishing different time scales for surface electromyography (EMG) signals, implanted EMG signals, and myophone signals. The coarse-grained sequence is given by the following formula: ,

[0052] Where N is the original time series The total length (i.e., the number of sample points); This represents the value of the j-th coarsening point at time scale τ; This represents the complete coarse-grained sequence corresponding to scale τ; then for each coarse-grained sequence... Calculate the sample entropy.

[0053] S300. Calculate the correlation coefficient between the first feature information and the second feature information.

[0054] For example, Figure 5 A flowchart for calculating the correlation coefficient between the first feature information and the second feature information in an embodiment of this application is shown below. Figure 5 As shown, the steps for calculating the correlation coefficient between the first feature information and the second feature information include S310-S330: S310. Calculate the covariance between the first feature information and the second feature information; S320. Determine the standard deviation of the first feature information and the standard deviation of the second feature information; S330. The correlation coefficient is calculated based on the covariance, the standard deviation of the first feature information, and the standard deviation of the second feature information.

[0055] In this embodiment of the application, the correlation coefficient between the first feature information and the second feature information is calculated. The formula is:

[0056] in, The covariance between the first feature information and the second feature information. The standard deviation of the first feature information. The standard deviation of the second feature information.

[0057] This application embodiment can also determine the correlation between surface electromyography signals and muscle sound signals by calculating the correlation coefficient between the first feature information and the third feature information. The formula for calculating the correlation coefficient between the first feature information and the third feature information is as follows:

[0058] in, The covariance between the first feature information and the third feature information. The standard deviation of the third feature information.

[0059] The correlation between implanted electromyography (EMG) signals and myophone signals can also be determined by calculating the correlation coefficient between the second and third feature information. The formula for calculating the correlation coefficient between the second and third feature information is as follows:

[0060] in, The covariance is the difference between the second and third feature information.

[0061] S400. Based on the correlation coefficient, determine whether there is a mapping relationship between the surface electromyography signal and the implanted electromyography signal.

[0062] For example, a statistical significance test value (also called a P-value) is determined between the first feature information and the second feature information. The statistical significance test value is used to characterize the probability that the difference between the first feature information and the second feature information is caused by random factors. If the absolute value of the correlation coefficient is detected to be greater than a first preset threshold and the statistical significance test value is less than a second preset threshold, a mapping relationship is determined between the surface electromyography (EMG) signal and the implanted EMG signal. A t-test can be used to first calculate the t-statistic, and the P-value can be obtained based on the t-statistic. The steps include: hypothesis setting: null hypothesis: there is no significant difference between the surface EMG signal and the implanted EMG signal in a certain dimension of characteristics; alternative hypothesis: there is a significant difference between the surface EMG signal and the implanted EMG signal in a certain dimension of characteristics. The t-statistic is calculated using the following formula:

[0063] in, The mean of the first feature information, The mean of the second feature information; The variance of the first feature information. The variance of the second feature information. The sample size for the first feature information. The sample size of the second feature information. Based on the calculated t-statistic, its corresponding p-value can be determined. The p-value is compared with a pre-set first significance level α (e.g., α=0.05) to make statistical inferences. If p<α, the null hypothesis is rejected and the alternative hypothesis is accepted, indicating that there is a statistically significant difference between the first and second feature information in the current dimension.

[0064] In this application example, by calculating the correlation between the multidimensional features of surface electromyography (EMG) signals and the multidimensional features of implanted EMG signals, it is possible to identify / prove that there is a significant linear correlation between surface EMG signals and implanted EMG signals, that is, there is a significant mapping relationship between the two.

[0065] S500. In the presence of the mapping relationship, verify that the third feature information of the myophone signal has an information compensation effect on the first feature information of the surface electromyography signal. In this embodiment, the information compensation effect can be understood as the characteristic information contained in the myophone signal that is related to the mechanical contraction activity of the muscle, which can provide the part of the deep muscle activity state that is not fully reflected in the surface electromyography signal characteristic information, thereby making up for the information loss when relying solely on the surface electromyography signal for inference.

[0066] For example, Figure 6 A flowchart illustrating the information compensation effect of the third feature information of the myophone signal on the first feature information of the surface electromyography signal provided in the embodiments of this application is shown below. Figure 6 As shown, the steps for verifying that the third characteristic information of the myophone signal has an information compensation effect on the first characteristic information of the surface electromyography signal include S510-S530: S510. Calculate the difference between the first feature information and the third feature information.

[0067] For example, the difference can be calculated through statistical hypothesis testing, which can employ a t-test, the steps of which include: (1) Hypothesis setting: Null hypothesis H0: There is no significant difference between myophone signal and surface electromyography signal in a certain dimension; Alternative hypothesis H1: There is a significant difference between myophone signal and surface electromyography signal in a certain dimension.

[0068] (2) Calculate the t-statistic, the formula is:

[0069] in, The mean of the first feature information, The mean of the third feature information; The variance of the first feature information. The variance of the third feature information. The sample size for the first feature information. The sample size of the third feature information.

[0070] (3) Based on the calculated t-value, determine its corresponding p-value (the p-value can be understood as the probability value caused by random factors used to assess the effect, difference, or correlation presented in the features between two related signals). Compare the p-value with the pre-set first significance level α (e.g., α=0.05) to make statistical inferences. If p<α, reject the null hypothesis H0 and accept the alternative hypothesis H1, indicating that there is a statistically significant difference between the first feature information and the third feature information in the current dimension. Further, when comparing different muscle activity states (e.g., normal contraction state and fatigue state), perform a t-test on the variance characteristics of specific channels of the surface electromyography signal. If the p-value is less than the pre-set second significance level (e.g., the pre-set second significance level is 0.001), it indicates that this feature of the surface electromyography signal has a highly significant statistical difference in different states, thus proving that the extracted features can effectively capture and distinguish different physiological states of muscle activity. This application embodiment also includes comparing various corresponding features (such as average energy in the time domain, median frequency in the frequency domain, nonlinear sample entropy, etc.) of surface electromyography (sEMG) and myosal sound (MMG) under the same muscle activity state. If the t-test results show that the p-value of the comparison between sEMG features and MMG features is preset at a second significance level in multiple feature dimensions, it indicates that the first feature information of surface electromyography and the third feature information of myosal sound have statistically significant differences in multiple aspects.

[0071] To further assess the practical significance of the difference (i.e., the magnitude of the difference), the effect size is calculated in addition to the statistical significance test. The calculation formula is as follows: ,in, This refers to the pooled standard deviation. If the effect size is greater than the preset threshold, it indicates that the difference is not only statistically significant but also practically significant.

[0072] S520. Construct a feature matrix from the first feature information, the second feature information, and the third feature information, and perform principal component analysis on the feature matrix to obtain the analysis results.

[0073] For example, before performing principal component analysis, the first feature information of the surface electromyography (EMG) signal, the second feature information of the implanted EMG signal, and the third feature information of the myophone signal are standardized to ensure the comparability of features of different dimensions. Further, the first, second, and third feature information are constructed into a feature matrix. Its dimensions are ,in, This refers to the number of samples (e.g., the number of time windows). The total number of features (e.g., 32 features) is represented by the matrix, where each row represents a sample and each column represents a feature.

[0074] For example, the steps of performing principal component analysis on the feature matrix to obtain the analysis results include S521-S527: S521. Calculate the covariance matrix of the feature matrix, and perform eigenvalue decomposition on the covariance matrix to obtain each eigenvalue and the corresponding eigenvector.

[0075] For example, the covariance matrix of the feature matrix is ​​calculated. The formula is:

[0076] The covariance matrix of this feature matrix can reflect the linear correlation between features.

[0077] The eigenvalue decomposition formula in this embodiment is:

[0078] in, It is a diagonal matrix containing eigenvalues. ; represents the eigenvector matrix, where each column is the eigenvector corresponding to the eigenvalue.

[0079] S522. Based on the eigenvalues, select k eigenvectors corresponding to the eigenvalues ​​to form the principal component space, where K is a positive integer greater than zero.

[0080] For example, in this embodiment of the application, the number of principal components to be retained, k, is determined by calculating the cumulative explained variance ratio. The formula for calculating the cumulative explained variance ratio is: In this application embodiment, the smallest variance is selected based on a preset explanatory variance ratio threshold. This means selecting the minimum number of principal components such that these principal components can jointly explain the variation information in the original data that is not lower than the threshold. In this embodiment, the eigenvectors corresponding to the first k eigenvalues ​​are selected to form the principal component space.

[0081] S523. Project the feature matrix onto the principal component space to obtain the dimensionality-reduced data matrix; For example, projecting the feature matrix into the principal component space yields a dimensionality-reduced data matrix, the formula of which is:

[0082] in, The data matrix after dimensionality reduction has a dimension of . ; For the front A matrix composed of eigenvectors.

[0083] S524. Calculate the first spatial distance between the first feature information and the second feature information, and the second spatial distance between the first feature information and the third feature information in the dimensionality-reduced data matrix; The first spatial distance refers to the distance between the center (mean vector) of the feature information sample points of the surface electromyography (EMG) signal and the center of the feature information sample points of the implanted EMG signal in the dimension-reduced data matrix. The second spatial distance refers to the distance between the center of the feature information sample points of all surface EMG signals and the center of the feature information sample points of the myocardial sound signal in the dimension-reduced data matrix, and its calculation formula is as follows: .in, , Let be the mean vectors of the two signal types in the PCA space. The first spatial distance and the second spatial distance can intuitively reflect the overall similarity or difference of the characteristics of different signal types in the dimensionality-reduced space.

[0084] S525. Obtain principal component loading information based on the feature vector; Here, loading information refers to the contribution of a certain feature to the principal components. For example, loading analysis is performed on the principal components to obtain loading information, and the formula is: ,in, Main component, For load information, representing features For principal components The contribution of the loading can be analyzed to identify the original features that contribute the most to a particular principal component.

[0085] S526. Calculate the contribution rate of each principal component based on the eigenvalues.

[0086] S527. The first spatial distance, the second spatial distance, the contribution rate, and the load information are used as the analysis results.

[0087] S530. Based on the differences and the analysis results, verify that the third feature information of the myophone signal has an information compensation effect on the first feature information of the surface electromyography signal.

[0088] For example, the step of verifying that the third feature information of the myophone signal has an information compensation effect on the first feature information of the surface electromyography signal includes S531-S535: S531. If the difference meets the preset conditions, determine that the first condition for verification is met; For example, based on the difference, the p-value and effect size between the first feature information and the third feature information are determined. If the difference meets a preset condition, it proves that the myophone signal provides information content different from the surface electromyography signal. The preset conditions include: the p-value of the difference is less than a third preset threshold and the effect size is greater than a fourth preset threshold. In this embodiment, if the p-value of the difference is less than the third preset threshold and the effect size is greater than the fourth preset threshold, the first verification condition is determined to be met.

[0089] S532. If the second spatial distance is greater than the first spatial distance, determine that the second verification condition is met; Specifically, the second spatial distance being greater than the first spatial distance indicates that, in the characteristic space, the overall distribution of the myophone signal differs from that of the surface electromyography (EMG) signal, which is greater than the difference between the surface EMG signal and its implanted EMG signal, indicating that the myophone signal and the surface EMG signal are complementary. In this embodiment, the second condition is: the second spatial distance is greater than the first spatial distance, and the difference between the first spatial distance and the second spatial distance is greater than a preset distance difference value.

[0090] S533. Based on the contribution rate and the load information, identify one or more principal components dominated by the third feature information, and calculate the total contribution rate of the one or more principal components.

[0091] The dominant criterion is that the absolute value of the load is greater than the preset load value, and the absolute value of the load of the third feature information in the same principal component is greater than the absolute value of the load of the first feature information and the second feature information; the total contribution rate refers to the sum of the contribution rates of all principal components identified that are dominated by the muscle sound signal features.

[0092] S534. If the total contribution rate is greater than a preset contribution rate threshold, the third verification condition is determined to be met. When the total contribution rate is greater than the preset contribution rate threshold, it indicates that the independent information carried by the myosalpingography signal constitutes an important and indispensable component of the overall characteristics of muscle activity.

[0093] S535. When the first condition, the second condition and the third condition are all satisfied, it is determined that the third feature information of the myophone signal has an information compensation effect on the first feature information of the surface electromyography signal.

[0094] In this embodiment, principal component analysis (PCA) can be used to obtain the overlap of features of surface electromyography (EMG), implanted EMG, and myophone signals in the principal component space, which can measure the degree of correlation between each pair of surface EMG, implanted EMG, and myophone signals. Through t-tests and PCA, it is verified that myophone signals can compensate for missing information in surface EMG signals, providing a scientific basis for subsequent mapping relationship identification.

[0095] In this embodiment of the application, to demonstrate the practical value of the information compensation function of myophone signals to surface electromyography (EMG) signals, a prediction model is constructed and a prediction experiment is conducted to evaluate the performance of the prediction model when performing prediction tasks (e.g., inferring implanted EMG signals from surface EMG signals) after fusing myophone signals, thereby proving the practical value of information compensation. Exemplarily, demonstrating the practical value of information compensation includes: 1. Prediction accuracy: Using implanted EMG signals as the prediction target, when only surface EMG signals are used as input to the prediction model, the average coefficient of determination R² of the prediction model is 0.78. When surface EMG signals and myophone signals are fused and used as input to the prediction model, the average R² of the prediction model is 0.92. Therefore, using surface EMG signals as input to the prediction model improves accuracy by 14%, proving the practical value of information compensation. 2. Residual Analysis: In the prediction model, the residuals predicted using only surface electromyography (SEMG) signals and the residuals predicted using both SEMG and myophone signals are calculated, and the variance of the residuals for both is calculated. The embodiments of this application show that the use of both SEMG and myophone signals is significantly reduced, indicating that myophone signals effectively compensate for the missing information in SEMG signals. 3. Calculation of Mutual Information between SEMG and Implantable EMG Signals: The SEMG and myophone signals are fused to obtain a fused signal. The mutual information between the fused signal and the implantable EMG signal is calculated. This result shows a significant increase in mutual information between the fused signal and the implantable EMG signal, confirming that the myophone signal provides important information missing from the SEMG signal.

[0096] Furthermore, to further verify and intuitively demonstrate the relationship between surface electromyography (EMG) signals, implanted EMG signals, and myophone signals from a nonlinear dimensionality reduction perspective, this application employs a t-distributed random neighborhood embedding algorithm for visualization analysis. This algorithm constructs and matches the probability distributions of data points in both the high-dimensional feature space and the low-dimensional visualization space, achieving a nonlinear preservation mapping of the high-dimensional structure. Subsequently, cluster analysis is used to explain the associations of different signal types. The steps include: Step 1: High-dimensional feature space similarity measurement: Construct a conditional probability distribution, the formula of which is:

[0097] in, , For samples in high-dimensional space, For Standard deviation of a Gaussian distribution centered at the center; symmetric probability .

[0098] To make the similarity measure symmetric, the perplexity is defined as: ,in, This is Shannon entropy. The perplexity in the embodiments of this application is used to control the size of the local neighborhood and balance the local and global structures.

[0099] Step 2, Low-dimensional embedding space mapping optimization: Define low-dimensional space similarity, its formula is:

[0100] in, , For mapping points in a low-dimensional (t-SNE) space, the t-distribution (degrees of freedom = 1, i.e., Cauchy distribution) is used to avoid the crowding problem.

[0101] We use KL divergence as the objective function to measure the difference between high-dimensional and low-dimensional distributions. The objective function is:

[0102] With the objective function as the goal, the location of low-dimensional points is optimized using gradient descent, and the gradient formula is:

[0103] For example, the optimization process can be set with an initial learning rate of 200, a total number of iterations of 1000, and a momentum parameter of 0.5 and an early compression strategy in the first 250 iterations to facilitate the stability of the initial layout, while the momentum parameter of 0.8 is used in subsequent iterations to accelerate convergence.

[0104] Step 3: Cluster Analysis: After obtaining the low-dimensional visualization mapping results, qualitative and quantitative analysis is performed on the clustering patterns of the data points to explain the relationships between different signal features. First, clusters are identified using DBSCAN or hierarchical clustering algorithms. After obtaining the clustering results, the quality of the clustering is evaluated using quantitative indicators such as the silhouette coefficient. The formula for calculating the silhouette coefficient is:

[0105] in, The average distance from a point to other points in the same cluster. The silhouette coefficient represents the average distance from a point to the nearest other cluster point. It should be noted that a larger silhouette coefficient indicates a denser cluster and a clearer separation between clusters.

[0106] The embodiments of this application, by employing the t-distributed random neighborhood embedding algorithm, can verify that the electromyography (EMG) signals, implanted EMG signals, and myophone signals form tight clusters in a low-dimensional space, further verifying the mapping relationships between surface EMG signals and implanted EMG signals, implanted EMG signals and myophone signals, and surface EMG signals and myophone signals. S600. Based on the information compensation effect, optimize the mapping relationship between the surface electromyography signal and the implanted electromyography signal.

[0107] This application's embodiments demonstrate a significant linear correlation between surface electromyography (SEMG) and implanted electromyography (EMG) signals through correlation analysis. Furthermore, t-tests and principal component analysis prove that myophone signals can effectively compensate for missing information in SEMG signals, improving the accuracy of the mapping between SEMG and implanted EMG signals. Using a t-distributed random neighborhood embedding algorithm, the distribution of SEMG, implanted EMG, and myophone signals in high-dimensional space is demonstrated, further validating the mapping relationships between SEMG and implanted EMG, implanted EMG and myophone signals, and SEMG and myophone signals. This provides a scientific basis for subsequent long-term non-invasive monitoring of specific muscle activities, significantly reducing the invasiveness and discomfort of long-term EMG monitoring, improving patient compliance, and is particularly suitable for scenarios requiring long-term monitoring, such as rehabilitation training, chronic muscle disease monitoring, and EMG prosthetic control.

[0108] To explain in detail the principles of the technical solution of this application, the overall process of this application will be described below with reference to some specific embodiments. It is easy to understand that the following is an explanation of the technical principles of this application and should not be regarded as a limitation of this application.

[0109] In a specific embodiment, the process for identifying the relationship between surface electromyographic signals and implanted electromyographic signals in this application includes the following steps: S1. Acquire surface electromyography (EMG) signals, implanted myocardial sound signals, and myocardial sound signals: Exemplarily, this application uses bullfrogs as experimental subjects and implants electrodes into them. The implantation point is determined based on the anatomical location of the target muscle. A radiofrequency ablation device is connected to the flexible electrode wire to ensure minimal muscle damage. The multi-channel flexible electrode wire is vertically inserted into the target muscle using radiofrequency ablation, ensuring a certain length distribution of the electrode wire within the muscle and that the tip of the electrode wire is located within the muscle fiber group. During insertion, short-term recording is used to observe the electromyographic signal characteristics to ensure correct electrode placement. Simultaneously, a dry electrode wristband (containing a multi-channel sEMG electrode array) is worn on the bullfrog's thigh, with the electrode position corresponding to the target muscle area. Good contact between the sEMG electrodes and the skin is ensured to guarantee signal quality. A liquid metal stretchable strain sensor is wrapped around the base of the bullfrog's thigh. Finally, the implanted electrode wire, dry electrode wristband, and liquid metal stretchable strain sensor are connected to an external connector. After connection, electromyographic activity is collected, including surface electromyographic signals, muscle sound signals, and muscle electromyographic signals. Thermal stimulation can be used to stimulate receptors in the bullfrog's feet, triggering a reflex arc that results in leg retraction / kickback movements. This allows for the acquisition of the correspondence between surface electromyography (EMG) signals and muscle EMG signals under normal conditions. Then, abnormal muscle activity is induced, and the correspondence between surface EMG signals and muscle EMG signals under these abnormal conditions is recorded. Furthermore, the positional changes of key movement sites (hip, knee, and ankle joints) and the delay from stimulus to response can be recorded simultaneously for later stratified analysis of different states. During the data acquisition and recording process, the positional changes of key movement sites (hip, knee, and ankle joints) and the delay from stimulus to response can be recorded simultaneously for later stratified analysis of different states.

[0110] S2. Preprocess the acquired surface electromyography (EMG) signals, muscle sound signals, and muscle EMG signals: The experimental data were organized and standardized for storage: For each bullfrog and each recording, surface electromyography (sEMG) signal matrices, myophonal signal matrices, and muscle EMG signal matrices (in the form of channel number × time point) were saved, along with metadata such as sampling rate, channel name, motion state, and abnormal state markers. Then, a uniform preprocessing procedure was applied to all channels: bandpass filtering (e.g., 10-600Hz) was performed on the sEMG and EMG channels, and a 50Hz and its harmonic notch filter was added to remove power frequency noise. After filtering, the acquired signals for each channel were mean-triggered and normalized according to their standard deviation to reduce the influence of individual differences and electrode contact differences on amplitude, making the mapping model more focused on temporal morphology and relative changes. Secondly, after filtering, artifact detection algorithms (e.g., based on amplitude, instantaneous energy, or derivative thresholds) were used to mark and remove time segments containing large-amplitude motion artifacts, saturation, or detachment, and to exclude or interpolate bad channels with long-term distortion.

[0111] S3. Extract the feature information corresponding to surface electromyography (EMG) signals, muscle sound signals, and muscle EMG signals: The feature information includes time-domain features, frequency-domain features, time-frequency features, and nonlinear dynamic system features. In this embodiment, the envelope of the effective signal is found through Hilbert transform and Gaussian smoothing filtering, extracting the following time-domain features: absolute value integral (reflecting signal strength), average energy (reflecting average signal strength), variance and standard deviation (reflecting signal fluctuation), and zero-crossing count (reflecting the speed of signal change). The power spectrum of the Fourier transform is calculated using a periodogram, extracting the following frequency-domain features: peak power corresponding frequency (reflecting the main frequency components), average power corresponding frequency (reflecting the average frequency components), and median power corresponding frequency (reflecting the median frequency components). The time-frequency characteristics of the signal are analyzed using wavelet transform (Daubechies wavelet), extracting features such as wavelet energy entropy. Based on chaos theory, the following nonlinear features are calculated: Lyapunov exponent (reflecting system stability), correlation dimension (reflecting system complexity), entropy and information entropy (reflecting system randomness), and complexity (comprehensively reflecting the system's complexity). S4. Identify whether there is a mapping relationship between surface electromyography (EMG) signals and implanted EMG signals: Calculate the correlation coefficient between the first feature information and the second feature information. The formula is:

[0112] in, The covariance between the first feature information and the second feature information. The standard deviation of the first feature information. The standard deviation of the second feature information.

[0113] This application embodiment can also determine the correlation between surface electromyography signals and muscle sound signals by calculating the correlation coefficient between the first feature information and the third feature information. The formula for calculating the correlation coefficient between the first feature information and the third feature information is as follows:

[0114] in, The covariance between the first feature information and the third feature information. The standard deviation of the third feature information.

[0115] The correlation between implanted electromyography (EMG) signals and myophone signals can also be determined by calculating the correlation coefficient between the second and third feature information. The formula for calculating the correlation coefficient between the second and third feature information is as follows:

[0116] in, The covariance is the difference between the second and third feature information.

[0117] Specifically, when the correlation coefficient between the first feature information and the second feature information is detected to be greater than a preset threshold, it is determined that there is a mapping relationship between the surface electromyography signal and the implanted electromyography signal.

[0118] S5. Verify that the third characteristic information of the myophone signal has an information compensation effect on the first characteristic information of the surface electromyography signal: 1. The t-test is used to calculate the t-statistic, and the formula is:

[0119] in, The mean of the first feature information, The mean of the third feature information; The variance of the first feature information. The variance of the third feature information. The sample size for the first feature information. The sample size of the third feature information.

[0120] The t-statistic was calculated to assess the significance of differences between surface electromyography (EMG) signals and implanted EMG signals, implanted EMG signals and myophone signals, and surface EMG signals and myophone signals.

[0121] 2. Principal component analysis (PCA) is performed on the first feature information of surface electromyography (EMG) signals, the second feature information of implanted EMG signals, and the third feature information of myophone signals. The high-dimensional feature space is projected into a low-dimensional space to show the feature distribution and correlation of surface EMG signals, implanted EMG signals, and myophone signals.

[0122] 3. The t-distributed random neighborhood embedding algorithm is used to map the high-dimensional feature space to a two-dimensional or three-dimensional space, visualize the clustering of sEMG, EMG, and MMG features, and interpret the correlation of different signal types through cluster analysis.

[0123] In summary, this application provides a method for identifying the relationship between surface electromyography (EMG) signals and implanted EMG signals of a muscle. This method first acquires surface EMG signals, implanted EMG signals, and myophone signals of the target muscle; then acquires first feature information of the surface EMG signals, second feature information of the implanted EMG signals, and third feature information of the myophone signals; calculates the correlation coefficient between the first and second feature information; and based on the correlation coefficient, determines whether a mapping relationship exists between the surface EMG signals and the implanted EMG signals. This method can identify whether a significant mapping relationship exists between the surface EMG signals and the implanted EMG signals, and objectively... The linear correlation between surface electromyography (EMG) signals and implanted EMG signals was quantitatively determined. In the presence of a mapping relationship, the third feature information of the myophone signal was verified to have an information compensation effect on the first feature information of the surface EMG signal, fully exploring the effective feature dimensions in the myophone signal that can compensate for the missing information in the surface EMG signal. Finally, based on the information compensation effect, the mapping relationship between surface EMG signals and implanted EMG signals was optimized, significantly improving the recognition accuracy of the mapping relationship between surface EMG signals and implanted EMG signals, making the identified mapping relationship more reliable, and providing reliable technical support for long-term non-invasive EMG monitoring.

[0124] like Figure 7 As shown in the diagram, this application also provides a structural schematic of a device for identifying the relationship between muscle surface electromyography (EMG) signals and implanted EMG signals, which can implement the above-mentioned method. This device may include: Signal acquisition module 21 is used to acquire surface electromyographic signals, implanted electromyographic signals and myophone signals of the target muscle; The feature information acquisition module 22 is used to acquire the first feature information of the surface electromyography signal, the second feature information of the implanted electromyography signal, and the third feature information of the myocardial sound signal; The correlation coefficient calculation module 23 is used to calculate the correlation coefficient between the first feature information and the second feature information; The relationship determination module 24 is used to determine, based on the correlation coefficient, whether there is a mapping relationship between the surface electromyography signal and the implanted electromyography signal; The information compensation verification module 25 is used to verify, in the presence of the mapping relationship, that the third feature information of the myophone signal has an information compensation effect on the first feature information of the surface electromyography signal; The relationship optimization module 26 is used to optimize the mapping relationship between the surface electromyography signal and the implanted electromyography signal based on the information compensation effect.

[0125] It is understood that the content of the above method embodiments is applicable to the present device embodiments. The specific functions implemented by the present device embodiments are the same as those of the above method embodiments, and the beneficial effects achieved are also the same as those achieved by the above method embodiments.

[0126] This application also provides an electronic device, which includes a memory and a processor. The memory stores a computer program, and the processor executes the computer program to implement the above-described method for identifying the relationship between surface electromyographic signals and implanted electromyographic signals. This electronic device can be any smart terminal, including tablet computers, in-vehicle computers, etc.

[0127] It is understood that the content of the above method embodiments is applicable to this device embodiment. The specific functions implemented by this device embodiment are the same as those of the above method embodiments, and the beneficial effects achieved are also the same as those achieved by the above method embodiments.

[0128] Please see Figure 8 , Figure 8 The hardware structure of an electronic device according to another embodiment is illustrated. The electronic device includes: The processor 901 can be implemented using a general-purpose CPU (Central Processing Unit), microprocessor, application-specific integrated circuit (ASIC), or one or more integrated circuits, and is used to execute relevant programs to implement the technical solutions provided in the embodiments of this application. The memory 902 can be implemented as a read-only memory (ROM), static storage device, dynamic storage device, or random access memory (RAM). The memory 902 can store the operating system and other application programs. When the technical solutions provided in the embodiments of this specification are implemented through software or firmware, the relevant program code is stored in the memory 902 and called by the processor 901 to execute the method for identifying the relationship between muscle surface electromyography signals and implanted electromyography signals according to the embodiments of this application. The input / output interface 903 is used to implement information input and output; The communication interface 904 is used to enable communication and interaction between this device and other devices. Communication can be achieved through wired means (such as USB, Ethernet cable, etc.) or wireless means (such as mobile network, WIFI, Bluetooth, etc.). Bus 905 transmits information between various components of the device (e.g., processor 901, memory 902, input / output interface 903, and communication interface 904); The processor 901, memory 902, input / output interface 903, and communication interface 904 are connected to each other within the device via bus 905.

[0129] This application also provides a computer-readable storage medium storing a computer program that, when executed by a processor, implements the above-described method for identifying the relationship between surface electromyographic signals and implanted electromyographic signals.

[0130] It is understood that the content of the above method embodiments is applicable to this storage medium embodiment. The specific functions implemented in this storage medium embodiment are the same as those in the above method embodiments, and the beneficial effects achieved are also the same as those achieved in the above method embodiments.

[0131] Memory, as a non-transitory computer-readable storage medium, can be used to store non-transitory software programs and non-transitory computer-executable programs. Furthermore, memory may include high-speed random access memory, and may also include non-transitory memory, such as at least one disk storage device, flash memory device, or other non-transitory solid-state storage device. In some embodiments, memory may optionally include memory remotely located relative to the processor, and these remote memories can be connected to the processor via a network. Examples of such networks include, but are not limited to, the Internet, intranets, local area networks, mobile communication networks, and combinations thereof.

[0132] The embodiments described in this application are for the purpose of more clearly illustrating the technical solutions of the embodiments of this application, and do not constitute a limitation on the technical solutions provided by the embodiments of this application. As those skilled in the art will know, with the evolution of technology and the emergence of new application scenarios, the technical solutions provided by the embodiments of this application are also applicable to similar technical problems.

[0133] Those skilled in the art will understand that the technical solutions shown in the figures do not constitute a limitation on the embodiments of this application, and may include more or fewer steps than shown, or combine certain steps, or different steps.

[0134] The device embodiments described above are merely illustrative. The units described as separate components may or may not be physically separate; that is, they may be located in one place or distributed across multiple network units. Some or all of the modules can be selected to achieve the purpose of this embodiment according to actual needs.

[0135] Those skilled in the art will understand that all or some of the steps in the methods disclosed above, as well as the functional modules / units in the systems and devices, can be implemented as software, firmware, hardware, or suitable combinations thereof.

[0136] In the several embodiments provided in this application, it should be understood that the disclosed apparatus and methods can be implemented in other ways. For example, the apparatus embodiments described above are merely illustrative; for instance, the division of the units described above is only a logical functional division, and in actual implementation, there may be other division methods. For example, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. Furthermore, the coupling or direct coupling or communication connection shown or discussed may be through some interfaces; the indirect coupling or communication connection between apparatuses or units may be electrical, mechanical, or other forms.

[0137] The units described above as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the units can be selected to achieve the purpose of this embodiment according to actual needs.

[0138] Furthermore, the functional units in the various embodiments of this application can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit. The integrated unit can be implemented in hardware or as a software functional unit.

[0139] If the integrated unit is implemented as a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of this application, in essence, or the part that contributes to the prior art, or all or part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes multiple instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods of the various embodiments of this application. The aforementioned storage medium includes various media capable of storing programs, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks. The preferred embodiments of the present application have been described above with reference to the accompanying drawings, but this does not limit the scope of the claims of the present application. Any modifications, equivalent substitutions, and improvements made by those skilled in the art without departing from the scope and substance of the embodiments of the present application shall be within the scope of the claims of the present application.

Claims

1. A method for identifying the relationship between surface electromyography (EMG) signals and implanted EMG signals, characterized in that, include: Acquire surface electromyographic signals, implanted electromyographic signals, and myophone signals of the target muscle; Acquire the first feature information of the surface electromyography signal, the second feature information of the implanted electromyography signal, and the third feature information of the myocardial sound signal; Calculate the correlation coefficient between the first feature information and the second feature information; Based on the correlation coefficient, determine whether there is a mapping relationship between the surface electromyography signal and the implanted electromyography signal; In the presence of the aforementioned mapping relationship, it is verified that the third feature information of the myophone signal has an information compensation effect on the first feature information of the surface electromyography signal; Based on the information compensation effect, the mapping relationship between the surface electromyography signal and the implanted electromyography signal is optimized.

2. The method for identifying the relationship between muscle surface electromyography (EMG) signals and implanted EMG signals according to claim 1, characterized in that, The first feature information, the second feature information, and the third feature information include time-domain features, frequency-domain features, time-frequency features, and nonlinear dynamic system features; The steps for obtaining the time-domain features, frequency-domain features, time-frequency features, and nonlinear dynamic system features include: The envelope information corresponding to the surface electromyography signal, implanted electromyography signal and myophone signal is obtained respectively, and the time-domain features in the first feature information, the time-domain features in the second feature information and the time-domain features in the third feature information are extracted from each corresponding envelope information. Calculate the power spectra corresponding to the surface electromyography signal, implanted electromyography signal, and myophone signal respectively, and extract the frequency domain features in the first feature information, the frequency domain features in the second feature information, and the frequency domain features in the third feature information from each corresponding power spectrum; Wavelet transforms are performed on the surface electromyography signal, implanted electromyography signal, and myophone signal, respectively. The time-frequency features in the first feature information, the time-frequency features in the second feature information, and the time-frequency features in the third feature information are extracted from the results of the corresponding wavelet transforms. The Lyapunov exponent, correlation dimension, entropy, information entropy, and complexity corresponding to surface electromyography (EMG), implanted EMG, and myosal signals are calculated respectively to obtain the nonlinear dynamic system features in the first feature information, the nonlinear dynamic system features in the second feature information, and the nonlinear dynamic system features in the third feature information.

3. The method for identifying the relationship between muscle surface electromyographic signals and implanted electromyographic signals according to claim 1, characterized in that, The calculation of the correlation coefficient between the first feature information and the second feature information includes: Calculate the covariance between the first feature information and the second feature information; Determine the standard deviation of the first feature information and the standard deviation of the second feature information; The correlation coefficient is calculated based on the covariance, the standard deviation of the first feature information, and the standard deviation of the second feature information.

4. The method for identifying the relationship between muscle surface electromyographic signals and implanted electromyographic signals according to claim 1, characterized in that, Determining whether a mapping relationship exists between the surface electromyography (EMG) signal and the implanted EMG signal based on the correlation coefficient includes: Determine the statistical significance test value between the first feature information and the second feature information, wherein the statistical significance test value is used to characterize the probability value that the difference between the first feature information and the second feature information is caused by random factors; If the absolute value of the correlation coefficient is greater than a first preset threshold and the statistical significance test value is less than a second preset threshold, it is determined that there is a mapping relationship between the surface electromyography signal and the implanted electromyography signal.

5. The method for identifying the relationship between muscle surface electromyography (EMG) signals and implanted EMG signals according to claim 1, characterized in that, The verification of the third characteristic information of the myophone signal has an information compensation effect on the first characteristic information of the surface electromyography signal, including: Calculate the difference between the first feature information and the third feature information; The first feature information, the second feature information, and the third feature information are used to construct a feature matrix. Principal component analysis is then performed on the feature matrix to obtain the analysis results. Based on the differences and the analysis results, it is verified that the third characteristic information of the myophone signal has an information compensation effect on the first characteristic information of the surface electromyography signal.

6. The method for identifying the relationship between muscle surface electromyography (EMG) signals and implanted EMG signals according to claim 5, characterized in that, The principal component analysis of the feature matrix yields the following results: Calculate the covariance matrix of the feature matrix, and perform eigenvalue decomposition on the covariance matrix to obtain each eigenvalue and the corresponding eigenvector; Based on the eigenvalues, select k eigenvectors corresponding to the eigenvalues ​​to form the principal component space, where K is a positive integer greater than zero; The feature matrix is ​​projected onto the principal component space to obtain the dimensionality-reduced data matrix; Calculate the first spatial distance between the first feature information and the second feature information, and the second spatial distance between the first feature information and the third feature information in the dimensionality-reduced data matrix; Principal component loading information is obtained based on the feature vector; The contribution rate of each principal component is calculated based on the aforementioned eigenvalues; The first spatial distance, the second spatial distance, the contribution rate, and the load information are used as the analysis results.

7. The method for identifying the relationship between muscle surface electromyography (EMG) signals and implanted EMG signals according to claim 6, characterized in that, The step of verifying, based on the differences and the analysis results, that the third characteristic information of the myophone signal has an information compensation effect on the first characteristic information of the surface electromyography signal includes: If the difference meets the preset conditions, it is determined that the first condition for verification is met; If the second spatial distance is greater than the first spatial distance, it is determined that the second verification condition is met; Based on the contribution rate and the load information, one or more principal components dominated by the third feature information are identified, and the total contribution rate of the one or more principal components is calculated. If the total contribution rate is greater than a preset contribution rate threshold, the third condition for verification is determined to be met. If the first condition, the second condition, and the third condition are all satisfied, it is determined that the third feature information of the myophone signal has an information compensation effect on the first feature information of the surface electromyography signal.

8. A device for identifying the relationship between surface electromyographic signals and implanted electromyographic signals, characterized in that, The device includes: The signal acquisition module is used to acquire surface electromyographic signals, implanted electromyographic signals, and myophone signals of the target muscle. The feature information acquisition module is used to acquire the first feature information of the surface electromyography signal, the second feature information of the implanted electromyography signal, and the third feature information of the myophone signal; The correlation coefficient calculation module is used to calculate the correlation coefficient between the first feature information and the second feature information. The relationship determination module is used to determine, based on the correlation coefficient, whether there is a mapping relationship between the surface electromyography signal and the implanted electromyography signal; The information compensation effect verification module is used to verify that, in the presence of the mapping relationship, the third feature information of the myophone signal has an information compensation effect on the first feature information of the surface electromyography signal. The relationship optimization module is used to optimize the mapping relationship between the surface electromyography signal and the implanted electromyography signal based on the information compensation effect.

9. An electronic device, characterized in that, Including the processor and memory; The memory is used to store programs; The processor executes the program to implement the method for identifying the relationship between muscle surface electromyography (EMG) signals and implanted EMG signals as described in any one of claims 1 to 7.

10. A computer program product, comprising a computer program, characterized in that, When the computer program is executed by the processor, it implements the method for identifying the relationship between surface electromyographic signals and implanted electromyographic signals as described in any one of claims 1 to 7.

Images